The deposit debate's blind spot

India's banking debate has increasingly narrowed to a single number: the credit-to-deposit (CD) ratio. But that number tells only a part of the story. CD ratio rose from 68.6 per cent in September 2021 to 82.2 per cent in March 2026, and has been read as evidence that banks have lent too much while failing to mobilise enough deposits.

The warning is simple: credit has run ahead of savings, and banks may hit a funding constraint. But that diagnosis does not fit the rest of the evidence. The banking system is healthier than it has been in years, with gross bad loans at 1.8 per cent, capital adequacy at 17.7 per cent and liquidity coverage well above the regulatory floor. The CD gap also looks less alarming in rupees than in growth rates. For instance, Rangarajan and Sambamurthy (businessline, August 19) noted the annual difference between credit and deposit accretion has mostly been modest, around ₹1-3 lakh crore in years when credit exceeded deposits, barely 1-1.5 per cent of the deposit base.

And when large foreign currency inflows recently swelled banking liquidity, the RBI drained nearly 60 per cent of the surplus within weeks. A system truly running out of funding does not usually coexist with liquidity absorption on this scale.

Savings in different form

The issue, therefore, is not that India has run out of savings. It is that banks are receiving savings in a different form. The old funding model relied on stable, granular, low-cost household deposits. The new one increasingly depends on money that reaches banks indirectly, in larger blocks, and with higher run-off risk.

That raises the marginal cost of funds and weakens the monetary transmission even when headline deposits look adequate. The RBI's latest rate hike to 5.50 per cent only sharpens this point: in a tightening cycle, the question is not whether deposits exist, but how quickly their cost reprices.

The simplest way to see the change is to stop asking whether deposits are growing and ask what kind of deposits they are. A household fixed deposit and a mutual fund's bank balance are both deposits. But they are not the same liability for a bank. One is sticky and predictable. The other is larger, more mobile, and reprices faster when returns change. In technical terms, the deposit beta rises, and funding volatility rises with it.

Chart 1 captures the shift. Deposits above ₹1 crore rose from 37 per cent of the total deposits in March 2020 to 44 per cent in March 2024. This does not mean households stopped saving. It means savings are returning to banks increasingly through institutional balance sheets.

This ownership shift is why the claim that stock markets are "draining deposits" is incomplete. At the system level, money does not vanish. It changes ownership. It moves into market products and returns as bulk deposits, which behave differently from retail money. The friction is not the existence of savings. It is the stability and pricing of bank liabilities.

This is the first break in the old model: savings are becoming market-led, while credit remains bank-led. The second break comes through the government's borrowing programme. Government borrowing does not create deposits when bonds are issued. Deposits return only when the government spends through salaries, transfers, subsidies, or contractor payments.

Fiscal cash management is, therefore, not just a treasury operation. It is a monetary variable. That would matter in any banking system, but it matters more today because the buyers of government securities have changed.

Chart 2 shows the shift. Non-bank institutions have overtaken commercial banks as the largest holders of Government of India (Gol) dated securities. When a bank buys government securities in primary issuance, it largely swaps one asset for another. But when a mutual fund, insurer, or pension fund funds the same borrowing, it pays from a bank deposit, and that deposit leaves the banking system until the government spends the proceeds.

What does this imply? If the state borrows early and spends late, banks compete for money that has temporarily moved out of their balance sheets. Paying more may help one bank win deposits from another, but it cannot create stable household deposits for the system. The result is a higher funding premium, not a larger funding base.

The same logic extends to the RBI balance sheet. Slower deposit growth can reflect weaker foreign asset accretion, lower net RBI credit to government and larger government cash balances at the central bank, all of which affect reserve money.

When the RBI runs a tighter, more hawkish liquidity stance, surplus is absorbed more aggressively; when it is more dovish, liquidity is allowed to persist. Either way, these are system flows, not branch-level failures.

The external account completes the story. Export receipts create deposits when dollars are converted into rupees. Import payments do the opposite. A current account deficit, therefore, pressures bank deposits as well as the rupee. When foreign inflows surge, deposit comfort can improve briefly, sometimes helped by rate-sensitive dynamics. The RBI may then sterilise part of the surplus to maintain its operating stance. Useful liquidity management, yes. A domestic funding base, no.

A stable option?

The policy question should, therefore, change from "are deposits growing fast enough?" to "what kind of deposits are banks getting?" Deposit mix, run-off risk, maturity mismatch and reliance on market funding matter more than the headline CD ratio.

Put simply, the deposit debate is asking the wrong question. Savings exist. The issue is whether a bank-led credit system can remain stable and cheap when its funding base shifts from sticky household money toward bulk, market-intermediated, and sometimes externally sourced deposits. That is not a problem one ratio can diagnose. It is a transition the system must manage.

Bhaduri is a Professor at Madras School of Economics; Anand is a PhD Scholar, MSE, Chennai. Views are personal

Published on October 9, 2026